Okay, here is the abstract and summary for the provided transcript, following the requested format.
Abstract:
This webinar recording from the IEEE UFFC Latin America series features announcements regarding academic and research positions, followed by two presentations and a Q&A session moderated by Prof. Elizabeth Hillman. The first talk, by Dr. Stacia Marie Fletcher, details research investigating the synergistic effects of microbubble-mediated focused ultrasound (FUS) combined with radiation therapy (RT) for treating glioblastoma in an f98 rat model. Findings show enhanced therapeutic effects (necrosis, ceramide increase) and reduced tumor growth, particularly with lower RT doses (4 Gy), leading to a modest but significant survival improvement compared to controls. Ongoing work focuses on histological analysis of apoptosis, ceramide, and immune markers. The second talk, by Danae Valenzuela Rodriguez, discusses the practical application and future potential of Artificial Intelligence (AI) in echocardiography. It highlights how AI is moving beyond hype to improve efficiency, reproducibility, and standardization through automated image classification, measurements (including strain), 3D visualization, and workflow optimization (stacked reading), while acknowledging challenges like user dependency, data quality, and the need for transparency and trust. The session concludes with a Q&A addressing tumor model limitations, FUS-induced immune responses, challenges and future directions for AI in both therapeutic and diagnostic ultrasound, including cross-modality integration.
AI and Focused Ultrasound in Medical Applications: Webinar Highlights
0:00:26 Introduction & Announcements: The webinar, part of the IEEE UFFC Latin America series, begins with announcements for PhD and academic/scientist positions in acoustical imaging and medical physics in the Netherlands, Chile, and Colombia.
0:01:53 Webinar Format: The session includes two talks (early career and senior speaker) followed by a combined Q&A session. Questions can be asked in English, Portuguese, or Spanish.
0:02:29 Moderator Introduction: Prof. Elizabeth Hillman from Columbia University is introduced as the moderator.
0:03:19 Speaker 1 Intro: Stacia Marie Fletcher, a postdoctoral fellow at Brigham and Women's/Harvard, is introduced. Her research focuses on ultrasound-enhanced drug delivery and non-invasive neurosurgery.
0:04:32 Talk 1 Start: FUS + Radiation Therapy Synergy: Dr. Fletcher discusses investigating the synergistic effects of microbubble-mediated focused ultrasound (FUS) and radiation therapy (RT) in an f98 glioma model.
0:05:06 Microbubble Bioeffects: Explanation of how microbubbles oscillate in an ultrasound field (cavitation) causing vascular bioeffects like mechanical stress and microstreaming, exploited for therapies like BBB opening.
0:06:32 Prior Work: Previous studies (e.g., Dr. Czarnota's group) showed FUS+microbubbles enhance RT effects in prostate, bladder, and breast tumors, increasing apoptosis and necrosis.
0:07:49 Mechanism involving Ceramide: RT can induce apoptosis not just via DNA damage but also via cell membrane stress, activating the ASMase pathway to produce ceramide, a secondary messenger initiating apoptosis. FUS+MB was shown to increase ceramide levels.
0:09:06 Study Goal: To investigate if similar synergistic effects occur in difficult-to-treat brain tumors (glioma).
0:09:37 Methods: A clinical FUS system (220 kHz) adapted for rats, defined ultrasound parameters, and single radiation doses (4, 8, 15 Gy) delivered via a small animal radiation platform.
0:10:13 Healthy Brain Results: Combined FUS+RT (15 Gy) caused more significant morphological changes (MRI, histology necrosis) and increased ceramide compared to FUS alone.
0:11:20 Tolerance: Long-term (21 days), FUS+RT (8 & 15 Gy) created persistent lesions, while FUS alone caused edema that generally resolved.
0:12:07 Tumor Model Results: f98 glioma cells implanted.
0:12:41 4 Gy RT: RT alone had little effect; FUS+RT significantly reduced tumor burden and increased tumor doubling time.
0:13:28 8 Gy RT: Highly variable response to RT alone; no significant difference found when adding FUS.
0:13:49 15 Gy RT: Strong response to RT alone; FUS+RT showed some further reduction but was not statistically significant in doubling time.
0:14:26 Survival Analysis: FUS+RT at 4 Gy resulted in a statistically significant survival increase compared to the control group, which was not seen with 4 Gy RT alone. The overall survival benefit was modest (~6% increase).
0:15:36 Talk 1 Conclusion: FUS+RT shows a moderate beneficial effect, suggesting potential but requires optimization. Ongoing work includes histological analysis of apoptosis, ceramide, and immune cell markers (like T-cells) to understand mechanisms and explore combination with immunotherapy.
0:17:10 Speaker 2 Intro: Danae Valenzuela Rodriguez, Head of Global Clinical Marketing at Philips Healthcare with extensive experience in echocardiography across multiple companies, is introduced.
0:18:46 Talk 2 Start: AI in Echocardiography: Ms. Rodriguez discusses the current state and potential of AI in echocardiography, framing it as assistive technology rather than a replacement.
0:19:48 AI Today: AI aims to increase efficiency, reproducibility, and improve outcomes in healthcare, moving from hype to practical applications enabled by better computing power.
0:22:17 AI Applications in Ultrasound: Includes computer-aided detection, decision support, and visualization enhancements. Examples: realistic 3D lighting, automated fetal biometry, automated 3D heart model quantification (volumes, EF, strain) in seconds.
0:26:01 Image Guidance & Fusion: AI enables fusing ultrasound with other modalities (like X-ray angiography in cath labs) for improved guidance during minimally invasive procedures, and advanced flow visualization (vessel flow, stent analysis).
0:29:01 AI Development Focus: Emphasis that significant effort goes into collecting and training data, not just coding algorithms.
0:30:00 Key AI Roles in Echo Now: Image classification (view recognition), automated measurements, and automated findings for decision support.
0:30:42 Addressing Variability: Ultrasound's user-dependency is a challenge; AI helps by standardizing view recognition and measurements.
0:32:15 Automated Workflows ("Stacked Reading"): AI recognizes imaging planes and automatically launches relevant measurements according to protocols, saving significant time and improving standardization.
0:34:10 Advanced Quantification (Strain): AI enables robust, automated analysis of myocardial deformation (strain), which is more reproducible and consistent than manual methods.
0:35:38 Benefits of AI in Echo: Minimizes user interaction, increases efficiency, reduces costs, maximizes reproducibility, improves diagnostic quality, and enables value-based care. Significant time savings demonstrated.
0:38:03 Addressing Healthcare Challenges: AI can help address physician shortages and manage data overload by automating routine tasks.
0:38:55 Future Vision: Integration of multi-modal data (ECG, US, MR, CT, labs) analyzed by AI for comprehensive patient assessment, prognosis, and treatment guidance.
0:39:44 AI Limitations & Human Role: AI assists, but physicians remain in control. Challenges include data dependency, algorithm transparency/explainability, and trust. AI excels at specific, trained tasks.
0:41:27 Talk 2 Conclusion: AI won't replace echocardiographers, but echocardiographers who utilize AI effectively may have an advantage over those who do not.
0:41:52 Q&A Start: Session opens for questions for both speakers.
0:42:28 Q&A (Stacia): Tumor Model Aggressiveness: Acknowledges f98 is aggressive and not ideal (avascular, non-infiltrative). Future work planned with more comparable human-derived xenografts.
0:43:46 Q&A (Stacia): Immune Response: Ongoing IHC analysis shows increased innate and adaptive immune cells with FUS and FUS+RT; the specific difference between FUS alone and FUS+RT is still under analysis.
0:45:46 Q&A (Danae): AI for Therapeutic Ultrasound (HIFU/Contrast): Notes challenges for AI in contrast echo due to regulatory variations and difficulty quantifying contrast uptake robustly (user/patient variability). Standardization is needed first before widespread AI analysis.
0:48:35 Q&A (Danae): AI Handling Ultrasound Variability/Noise: Acknowledges US is challenging for AI due to low SNR and high user/patient variability. Current AI focus is on standardizing routine tasks (view recognition, measurements) which addresses ~80% of daily work, providing significant efficiency gains as a first step.
0:53:44 Q&A (Stacia): Radiation Dose Response (8 vs 15 Gy): 15 Gy likely ablative; 8 Gy is on a threshold causing variable biological responses. High single doses may trigger different cell death pathways (e.g., membrane/ceramide) leading to immediate effects not typical of fractionated RT.
0:57:07 Q&A (Danae): Future AI in Echo Systems: Key directions include fully automated "stacked reading" protocols, cross-modality data integration (combining US, MR, CT data for prognosis, e.g., cardio-oncology), and improved fusion imaging for procedural guidance (breaking down modality silos).
1:00:50 Conclusion & Next Webinar: Thanks to speakers and moderator; announcement of the next webinar in February 2023.